Version 1
: Received: 12 May 2024 / Approved: 12 May 2024 / Online: 13 May 2024 (12:17:46 CEST)
How to cite:
Liu, A.; Sabhanayakam, K. Mitigating Overfitting in Neural Net Classification of Radio Galaxies through Wavelet Analysis. Preprints2024, 2024050800. https://doi.org/10.20944/preprints202405.0800.v1
Liu, A.; Sabhanayakam, K. Mitigating Overfitting in Neural Net Classification of Radio Galaxies through Wavelet Analysis. Preprints 2024, 2024050800. https://doi.org/10.20944/preprints202405.0800.v1
Liu, A.; Sabhanayakam, K. Mitigating Overfitting in Neural Net Classification of Radio Galaxies through Wavelet Analysis. Preprints2024, 2024050800. https://doi.org/10.20944/preprints202405.0800.v1
APA Style
Liu, A., & Sabhanayakam, K. (2024). Mitigating Overfitting in Neural Net Classification of Radio Galaxies through Wavelet Analysis. Preprints. https://doi.org/10.20944/preprints202405.0800.v1
Chicago/Turabian Style
Liu, A. and Karthik Sabhanayakam. 2024 "Mitigating Overfitting in Neural Net Classification of Radio Galaxies through Wavelet Analysis" Preprints. https://doi.org/10.20944/preprints202405.0800.v1
Abstract
Over the last few decades, advancements in astrophysics have been closely linked to the developmentof powerful machine-learning models that can accurately classify celestial bodies. At the same time,however, many astronomical datasets are filled with new features collected by increasingly powerfultelescopes. These features can cause overfitting, clouding predictive abilities, and dampening theability of many models to classify images. Therefore, motivation to design more efficient models hasskyrocketed—aiming to optimize for lower run times and high accuracies, even with fewer providedfeatures. In our project, we seek to optimize a convolutional neural network model using a techniqueknown as wavelet analysis. This technique allows us to pick the key features of an astronomicalimage and accentuate niche details, saving time and boosting accuracy. We applied it to the MiraBestDataset, a dataset compiled from the FIRST sky survey using a virtual telescope. In the end, aftertraining our neural network on the original images and the five filters (approximation, horizontal,vertical, diagonal, and combined), we found that with fewer features and less overfitting, the verticalDaubechies-family wavelet filter outperformed the original runs with the unaltered images by over10%. Our findings suggest that wavelet analysis can help harvest the most valuable features inimages of celestial bodies–leading to enhanced predictions in astronomical applications and perhapsbolstering modern astrophysical theory.
Keywords
Machine Learning; Astrophysics; Computer Vision
Subject
Computer Science and Mathematics, Computer Vision and Graphics
Copyright:
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.